The applicants proposed to develop high resolution detectors for use in a dedicated PET system for imaging small laboratory animals ("microPET"). This device would have many applications, including biodistribution studies of labeled drugs and new tracers, studies of organ function, imaging of gene expression and monitoring the effects of therapy on tumors. High resolution brain studies in cats and primates would also lead to exciting opportunities in neuroscience. Previous attempts at building dedicated animal scanners have used the same block detector technology found in commercial PET systems. The detectors in this application are designed to overcome the limitations of the block design, at the same time satisfying the sensitivity and packing fraction demands of PET imaging. The detectors are based on an array of individual 2x2xl0 mm BGO crystals, coupled via optical fibers to a 64 element multichannel photomultiplier tube. The one to one coupling between scintillator and photodetector element eliminates the use of light sharing for position determination, leading to high spatial resolution and the possibility of rejecting events which scatter in the detector. Two detector modules have been built and exhibit a spatial resolution of 1.4 mm. Detailed preliminary data is presented to further support the feasibility of our approach. The applicants described their plans to optimize the performance of these detectors and to build a small, low cost, prototype system for evaluation. This system would consist of a partial ring of detector modules mounted on an accurate rotation table allowing tomographic datasets to be acquired. They also described the experiments (phantom and in vivo animal studies) to be performed to assess the performance of the prototype system. The applicants also proposed to investigate data correction methods, different data sampling schemes, 3-D image reconstruction and resolution recovery techniques. Simulations have shown that the prototype microPET system should attain a reconstructed spatial resolution of <2 mm in each dimension, leading to a volume resolution element of under 8 mm3. This is a six-fold improvement over the best PET systems currently in existence. Although the resolution is still coarse relative to autoradiography, it does offer an important advantage in that kinetic information can be obtained from a single animal. In addition, repeat studies can be performed on the same animal before and after pharmacological or surgical intervention. It is therefore expected that the ability of the microPET system to noninvasively image and quantify functional changes within a single animal over a time course of minutes, hours, days or months will ultimately provide important new information to the basic biomedical sciences.